public static void Go(object exempl1)
{
Exempl exempl = (Exempl)exempl1;
for (double i = exempl.start; i < exempl.end; i += 0.1)
{
for (int j = 2; j < 7; j++)
{
Topology topology = new Topology(6, 4, i, j);
NeuralNetwork neuralNetwork = new NeuralNetwork(topology);
var diff1 = neuralNetwork.Learn(exempl.dataset, exempl.dataset2, size);
mutex.WaitOne();
if (diff1[0].Last() /* * diff1[0].Last() + (diff1[1].Last() - diff1[0].Last()) * (diff1[1].Last() - diff1[0].Last()) */ < minDif)
{
step = i;
count = j;
minDif = diff1[0].Last() /** diff1[0].Last() +( diff1[1].Last() - diff1[0].Last())* (diff1[1].Last() - diff1[0].Last())*/;
}
mutex.ReleaseMutex();
}
}
}
static void Main(string[] args)
{
Bitmap bitmap = new Bitmap(Properties.Resources.Image1);
var dataset = new List <Tuple <double[], double[]> >
{
new Tuple <double[], double[]>(new double[] { 1, 0, 0, 0 }, new double[] { StrToDouble(0, 0, 1, 0, 0), StrToDouble(0, 1, 1, 0, 0), StrToDouble(0, 0, 1, 0, 0), StrToDouble(0, 0, 1, 0, 0), StrToDouble(0, 0, 1, 0, 0), StrToDouble(0, 1, 1, 1, 0) }),
new Tuple <double[], double[]>(new double[] { 0, 1, 0, 0 }, new double[] { StrToDouble(0, 1, 1, 1, 0), StrToDouble(0, 1, 0, 1, 0), StrToDouble(0, 0, 0, 1, 0), StrToDouble(0, 0, 1, 0, 0), StrToDouble(0, 1, 0, 0, 0), StrToDouble(0, 1, 1, 1, 0) }),
new Tuple <double[], double[]>(new double[] { 0, 0, 1, 0 }, new double[] { StrToDouble(1, 0, 0, 1, 0), StrToDouble(1, 0, 0, 1, 0), StrToDouble(1, 0, 0, 1, 0), StrToDouble(1, 0, 0, 1, 0), StrToDouble(1, 1, 1, 1, 1), StrToDouble(0, 0, 0, 0, 1) }),
new Tuple <double[], double[]>(new double[] { 0, 0, 0, 1 }, new double[] { StrToDouble(1, 0, 0, 1, 0), StrToDouble(1, 0, 1, 0, 1), StrToDouble(1, 1, 1, 0, 1), StrToDouble(1, 1, 1, 0, 1), StrToDouble(1, 0, 1, 0, 1), StrToDouble(1, 0, 0, 1, 0) }),
/*new Tuple<double[],double[]>(new double[]{1,0,0,0 },ImageToDouble(Properties.Resources.Image1)),
* new Tuple<double[],double[]>(new double[]{0,1,0,0 },ImageToDouble(Properties.Resources.Image2)),
* new Tuple<double[],double[]>(new double[]{0,0,1,0 },ImageToDouble(Properties.Resources.Image3)),
* new Tuple<double[],double[]>(new double[]{0,0,0,1 },ImageToDouble(Properties.Resources.Image4))
*/
};
var dataset2 = new List <Tuple <double[], double[]> >
{
new Tuple <double[], double[]>(new double[] { 1, 0, 0, 0 }, new double[] { StrToDouble(1, 0, 1, 0, 0), StrToDouble(0, 1, 1, 0, 1), StrToDouble(0, 0, 1, 0, 0), StrToDouble(0, 0, 0, 0, 0), StrToDouble(0, 0, 1, 1, 0), StrToDouble(0, 1, 1, 0, 0) }),
new Tuple <double[], double[]>(new double[] { 0, 1, 0, 0 }, new double[] { StrToDouble(0, 0, 1, 1, 0), StrToDouble(1, 1, 0, 1, 0), StrToDouble(1, 0, 0, 1, 0), StrToDouble(0, 0, 1, 1, 0), StrToDouble(0, 1, 0, 0, 1), StrToDouble(0, 0, 1, 1, 0) }),
new Tuple <double[], double[]>(new double[] { 0, 0, 1, 0 }, new double[] { StrToDouble(1, 0, 1, 1, 0), StrToDouble(1, 0, 0, 1, 0), StrToDouble(0, 0, 0, 1, 0), StrToDouble(1, 0, 0, 1, 0), StrToDouble(1, 1, 0, 1, 1), StrToDouble(1, 0, 0, 0, 1) }),
new Tuple <double[], double[]>(new double[] { 0, 0, 0, 1 }, new double[] { StrToDouble(1, 0, 0, 1, 1), StrToDouble(1, 0, 1, 1, 1), StrToDouble(1, 1, 0, 0, 1), StrToDouble(1, 1, 1, 1, 1), StrToDouble(1, 1, 1, 0, 1), StrToDouble(0, 0, 0, 1, 0) }),
/*new Tuple<double[],double[]>(new double[]{1,0,0,0 },ImageToDouble(Properties.Resources.Image1)),
* new Tuple<double[],double[]>(new double[]{0,1,0,0 },ImageToDouble(Properties.Resources.Image2)),
* new Tuple<double[],double[]>(new double[]{0,0,1,0 },ImageToDouble(Properties.Resources.Image3)),
* new Tuple<double[],double[]>(new double[]{0,0,0,1 },ImageToDouble(Properties.Resources.Image4))
*/
};
var ex = new Exempl[4];
double start = 2.0;
double end = 5.0;
double st = (end - start) / ex.Length;
for (int i = 0; i < ex.Length; i++)
{
ex[i] = new Exempl(start, start + st, dataset, dataset2);
start += st;
}
Thread[] threads = new Thread[4];
for (int i = 0; i < threads.Length; i++)
{
threads[i] = new Thread(new ParameterizedThreadStart(Go));
}
for (int i = 0; i < threads.Length; i++)
{
threads[i].Start(ex[i]);
}
for (int i = 0; i < threads.Length; i++)
{
threads[i].Join();
}
NeuralNetwork neuralNetwork;
Topology topology;
topology = new Topology(6, 4, step, count);
neuralNetwork = new NeuralNetwork(topology);
var diff = neuralNetwork.Learn(dataset, dataset2, size);
//-------------------------------------
MLApp.MLApp matlab = new MLApp.MLApp();
object result;
double[] mas = new double[size];
for (int i = 0; i < size; i++)
{
mas[i] = i;
}
matlab.Execute(@"cd C:\Users\Arseniy\source\repos\NeuronNetworks\NeuronNetworks");
matlab.Feval("myfunc", 1, out result, mas, diff[0], diff[1]);
foreach (var data in dataset)
{
neuralNetwork.FeedForward(data.Item2);
}
foreach (var data in dataset2)
{
neuralNetwork.FeedForward(data.Item2);
}
}
